Intelligent Crop Recommendation through Multi-Modal Deep Learning and Remote Sensing Analytics

Abstract

Accurate crop selection requires not only an understanding of environmental suitability but also consideration of future market conditions and economic returns. This study proposes an intelligent agricultural decision-support framework that integrates crop yield prediction, demand forecasting, price prediction, market intelligence, and crop recommendation within a unified platform. Agricultural data obtained from the Government of India’s Open Government Data (OGD) platform were combined with environmental parameters such as rainfall, temperature, and soil fertility to support predictive analytics and recommendation generation. The proposed framework performs data preprocessing, feature engineering, demand forecasting, yield and price prediction, demand–supply analysis, and profit-oriented crop recommendation. A web-based dashboard was developed to provide location-specific agricultural insights and decision support. Experimental results demonstrate the effectiveness of the framework in forecasting demand over multiple time horizons, predicting crop yield and market prices, identifying future supply deficits and surpluses, and recommending crops with higher economic potential. The findings reveal that integrating market intelligence with agronomic prediction enhances agricultural decision-making, supports sustainable crop planning, and improves farmer profitability. The proposed system represents a significant step toward precision agriculture by combining productivity assessment and market-driven analytics to deliver intelligent, data-driven, and profit-oriented crop recommendations